Elastic Cloud Storage (ECS) uses the latest trends in software architecture and development to achieve increased availability, capacity use efficiency, and performance. ECS uses a specific method for disk capacity management. Disk space is partitioned into a set of blocks of fixed size called chunks. User data is stored in these chunks and the chunks are shared. One chunk can contain fragments of several user objects. Chunk content is modified in an append mode. When chunks become full, they are sealed and the content of sealed chunks is immutable.
A corresponding feature of ECS supports geographically distributed setups (GEO) consisting of two or more zones. When there are three or more zones, each zone may receive chunks from other zones and the zones can combine the chunks. This type of cluster based storage of several racks supports cluster nodes that communicate with each other. One cluster may be located in one data center and another cluster can be located in a remote data center. Although the geographically distributed setups can be used to provide an additional protection of user data by means of replication, the replication mechanism works at the chunks level. Therefore, this presents a scalability issue because there is no way for the geographically distributed setup to help to recover data from a dual failure or from more complicated failures.
The above-described background relating to ECS is merely intended to provide a contextual overview of some current issues, and is not intended to be exhaustive. Other contextual information may become further apparent upon review of the following detailed description.